Schematic representation of structural parameters of healthy (A) and tumor (B) blood vessels. A, In healthy tissue, blood vessels develop a well-organized network of arteries and veins to support the need for oxygen and nutrients (upper part panel A), with a normal vessel wall and endothelium (lower part of panel A). B, In a rapidly growing tumor, both the vasculature (upper part of panel B), as well as the vessel wall and endothelium (lower part of panel B) are abnormal in shape and structure, resulting in increased hypoxia (represented by blue shading) and IFP. BM, basement membrane; ECs, endothelial cells; IFP, interstitial fluid pressure. Adapted from (Carmeliet and Jain, 2011b).
In addition to its crucial role as a regulator of sprouting angiogenesis, VEGF plays a critical role in promoting the permeability of tumor blood vessels, thereby further facilitating the intravasation of cancer cells (Dvorak et al., 1999). Moreover, hypoxia-induced ANG-2 destabilizes tumor blood vessels (reduces the coverage of ECs by pericytes) by antagonizing the activity of anigopoetin-1 (ANG-1) (Falcon et al., 2009).
Because of the fundamental role of angiogenesis in cancer, blocking molecules involved in angiogenesis appeared to be an attractive strategy for cancer therapies
3.1.1. ANTI-ANGIOGENIC THERAPY: SUCCESSES AND LIMITATION
Since the concept of targeting angiogenesis to starve tumors to death was introduced, more than 40 years ago (Folkman, 1971), ten anti-angiogenic drugs targeting VEGF or its receptors are approved for cancer therapy, with many more in clinical trails (Jain, 2014). Among these, the pioneers in angiogenesis inhibitors are: a ligand-trapping humanized monoclonal antibody against VEGF-A – bevacizumab (Avastin, Genentech / Roche), and two kinase inhibitors – sorafenib (Nexavar, Bayer) and sunitinib (Sutent, Pfizer) – that target tyrosine kinases (among others VEGFR-2).
Despite the encouraging intial results in pre-clinical models and clinical trials, anti-angiogenic therapy by itself failed to meet the expectations. These initial benefits, in the order of weeks to months (at best), are only transitory, because of either intrinsic resistance of tumors to these agents prior to treatment or the development of resistance after an initial response (Bergers and Hanahan, 2008; Jain, 2014). In the latter case, the evasion of anti-angiogenic therapy is mediated by the development of alternative mechanisms to sustain tumor growth.
In part, the following adaptive can be involved: (i) activation and / or upregulation of alternative pro-angiogenic signaling pathways; (ii) activation of vasculogenesis; (iii) increased pericyte coverage of the tumor vessels, which confers resistance to therapy; (iv) increased invasion of cancer cells that allows them to access the blood supply of normal tissue without the necessity of neovascularization (Bergers and Hanahan, 2008).
Moreover, VEGF inhibition as a single therapy has been suggested to promote a switch of the cancer cells to a more invasive and more metastatic phenotype, possibly contributing to the limitations of patients’ survival (Kerbel, 2008). Indeed, the results from pre-clinical studies on several tumor models confirmed that inhibition of angiogenesis as a monotherapy promotes tumor progression to malignancy, with heightened local invasion and increased metastatic capacity (Ebos et al., 2009; Paez-Ribes et al., 2009).
On the contrary to these ‘vessel pruning’ strategies, recent preclinical and initial clinical results revealed that ‘normalization of the vascular abnormalities’ in cancer is emerging as a complementary therapeutic approach to standardized cancer therapies (Carmeliet and Jain, 2011b; Jain, 2014).
3.1.2. VESSEL NORMALIZATION AS A NEW THERAPEUTHIC APPROACH TO BLOCK METASTASIS
In tumors, the production of angiogenic factors is excessive and leads to the development of functionally and structurally abnormal vessels. The strategy of tumor vessel normalization aims to restore the equilibrium of pro- and anti- angiogenic factors that are skewed in the tumor microenvironment. This would revert the utterly abnormal structure and function of the tumor vessels towards a more normal state (Goel et al., 2011). These changes would not only contribute
properties, but could also contribute to an improved chemoresponse by improving drug delivery.
The importance of vessel normalization has been recognized in 2001. The idea stemmed from the observation that monotherapy of bevacizumab had divergent outcomes compared to combined bevacizumab and chemotherapy on overall survival in patients (Jain, 2001). While monotherapy with an anti-VEGF monoclonal antibody failed to prolong the overall survival of patients, combined treatment with chemotherapy conferred a survival benefit (Jain, 2001; Jain, 2014). This beneficial effect was mediated by transient ‘normalization’ of the abnormal vasculature of tumors as a result of anti-angiogenic treatment that temporally improved tumor perfusion and reduced intratumoral hypoxia (known to promote resistance to chemotheraphy and radiotherapy). Thus, therapies that were administered during the window of time when tumor vessels were normalized might have achieved better efficacy (Jain, 2001; Jain, 2014).
Since 2001, several preclinical studies using direct and indirect anti- angiogenic agents supported the tumor vessel normalization hypothesis (Izumi et al., 2002; Tong et al., 2004; Winkler et al., 2004; Yuan et al., 1996). Moreover, a number of additional targets, present either in cancer cells or in stromal cells, have been implicated in facilitating or hindering vessel normalization (Jain, 2014). Among these, studies from our lab showed that genetic blockade of PHD2 in ECs promotes tumor vessel normalization (Mazzone et al., 2009). Moreover, we uncovered unanticipated activities of chloroquine (FDA-approved antimalarial drug) as a tumor vessel normalizing agent (Maes et al., 2014).
Altogether, these studies support the notion that tumor vessel ‘normalization’, rather than vessel ‘pruning’, presents a promising strategy to diminish cancel cell invasive and metastatic capacity, whilst improving drug- delivery and chemotherapy-response, thus promising a prolonged overall survival.
3.2. CANCER-ASSOCIATED FIBROBLASTS
CAFs are the most prominent cell type within the tumor stroma of many tumors, especially in breast, prostate and pancreatic carcinomas (Kalluri and Zeisberg, 2006; Pietras and Ostman, 2010). In fact, the fibroblast population in pancreatic cancers may comprise more than 90% of the overall tumor mass (Karagiannis et al., 2012).
The transformation of local fibroblasts is pathologically important in the progression of cancer. CAFs are a form of perpetually activated fibroblasts. They proliferate faster and deposit higher amounts of ECM components than resting fibroblasts in benign tissue (Kalluri and Zeisberg, 2006). While in the early phase of carcinogenesis fibroblasts can have tumor suppressing activities, the phenotype of activated fibroblasts (CAFs) switches to a tumor promoting state as the tumor progresses (Proia and Kuperwasser, 2005). CAFs have an important role in tumorigenesis and in malignant progression. They facilitate proliferation, invasion and motility of malignant cells, and eventually contribute to metastasis (Cirri and Chiarugi, 2011). Their pro-tumorigenic and pro-metastatic activities are related to several mechanisms (Figure 7).
First, CAFs facilitate tumor cell invasion through protease- and force- dependent matrix remodeling (Calvo et al., 2013; Gaggioli et al., 2007). They stimulate angiogenesis and provide building blocks (amino acids, nucleotides) and metabolites (glutamine, pyruvate, lactate) for cancer cell metabolism, promoting tumor growth, invasion and metastasis (Liu et al., 2011). CAFs and hypoxia are both crucial for tumor progression, however little is known about how hypoxia affects the recruitment and activation of CAFs. It has recently been indicated that the recruitment of fibrocytes (possible CAF precursors) / myofibroblasts to sites of pathological fibrosis may be driven by hypoxia (Giaccia and Schipani, 2010). Although it is unknown whether tumor hypoxia does indeed play a role in the recruitment of CAF precursors to the tumor, tumor hypoxia may
control the differentiation of CAFs from precursor cells through mechanisms involving TGF-β1 and endothelin-1 (ET-1) (Bellini and Mattoli, 2007). Overall, hypoxia might enhance most of the pro-tumorigenic activities of CAFs.